[prev in list] [next in list] [prev in thread] [next in thread] 

List:       llvm-bugs
Subject:    [llvm-bugs] [Bug 43515] New: trunc(x >> (bitwidth-1)) ^ 1 is not recognized as bt+setae
From:       via llvm-bugs <llvm-bugs () lists ! llvm ! org>
Date:       2019-09-30 21:51:55
Message-ID: bug-43515-206 () http ! bugs ! llvm ! org/
[Download RAW message or body]

--1569880315.8e3413.10997
Date: Mon, 30 Sep 2019 14:51:55 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://bugs.llvm.org/
Auto-Submitted: auto-generated

https://bugs.llvm.org/show_bug.cgi?id=43515

            Bug ID: 43515
           Summary: trunc(x >> (bitwidth-1)) ^ 1 is not recognized as
                    bt+setae
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Backend: X86
          Assignee: unassignedbugs@nondot.org
          Reporter: lebedev.ri@gmail.com
                CC: craig.topper@gmail.com, llvm-bugs@lists.llvm.org,
                    llvm-dev@redking.me.uk, spatel+llvm@rotateright.com

https://godbolt.org/z/jUiaJg

If we start with
int cmpge32_0(long a) {
  return a>=0;
}
We get
        shr     rax, 63
        xor     eax, 1

But for less canonical 
  %n0 = and i64 %x, 9223372036854775808 # sign bit mask
  %n1 = icmp eq i64 %n0, 0
  %r = zext i1 %n1 to i32 

we get seemingly-more-optimal
        xor     eax, eax
        bt      rdi, 63
        setae   al

-- 
You are receiving this mail because:
You are on the CC list for the bug.
--1569880315.8e3413.10997
Date: Mon, 30 Sep 2019 14:51:55 -0700
MIME-Version: 1.0
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://bugs.llvm.org/
Auto-Submitted: auto-generated

<html>
    <head>
      <base href="https://bugs.llvm.org/">
    </head>
    <body><table border="1" cellspacing="0" cellpadding="8">
        <tr>
          <th>Bug ID</th>
          <td><a class="bz_bug_link 
          bz_status_NEW "
   title="NEW - trunc(x &gt;&gt; (bitwidth-1)) ^ 1 is not recognized as bt+setae"
   href="https://bugs.llvm.org/show_bug.cgi?id=43515">43515</a>
          </td>
        </tr>

        <tr>
          <th>Summary</th>
          <td>trunc(x &gt;&gt; (bitwidth-1)) ^ 1 is not recognized as bt+setae
          </td>
        </tr>

        <tr>
          <th>Product</th>
          <td>libraries
          </td>
        </tr>

        <tr>
          <th>Version</th>
          <td>trunk
          </td>
        </tr>

        <tr>
          <th>Hardware</th>
          <td>PC
          </td>
        </tr>

        <tr>
          <th>OS</th>
          <td>Linux
          </td>
        </tr>

        <tr>
          <th>Status</th>
          <td>NEW
          </td>
        </tr>

        <tr>
          <th>Severity</th>
          <td>enhancement
          </td>
        </tr>

        <tr>
          <th>Priority</th>
          <td>P
          </td>
        </tr>

        <tr>
          <th>Component</th>
          <td>Backend: X86
          </td>
        </tr>

        <tr>
          <th>Assignee</th>
          <td>unassignedbugs&#64;nondot.org
          </td>
        </tr>

        <tr>
          <th>Reporter</th>
          <td>lebedev.ri&#64;gmail.com
          </td>
        </tr>

        <tr>
          <th>CC</th>
          <td>craig.topper&#64;gmail.com, llvm-bugs&#64;lists.llvm.org, \
llvm-dev&#64;redking.me.uk, spatel+llvm&#64;rotateright.com  </td>
        </tr></table>
      <p>
        <div>
        <pre><a href="https://godbolt.org/z/jUiaJg">https://godbolt.org/z/jUiaJg</a>

If we start with
int cmpge32_0(long a) {
  return a&gt;=0;
}
We get
        shr     rax, 63
        xor     eax, 1

But for less canonical 
  %n0 = and i64 %x, 9223372036854775808 # sign bit mask
  %n1 = icmp eq i64 %n0, 0
  %r = zext i1 %n1 to i32 

we get seemingly-more-optimal
        xor     eax, eax
        bt      rdi, 63
        setae   al</pre>
        </div>
      </p>


      <hr>
      <span>You are receiving this mail because:</span>

      <ul>
          <li>You are on the CC list for the bug.</li>
      </ul>
    </body>
</html>
--1569880315.8e3413.10997--


[Attachment #3 (text/plain)]

_______________________________________________
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs


[prev in list] [next in list] [prev in thread] [next in thread] 

Configure | About | News | Add a list | Sponsored by KoreLogic